Assigning time-linked exposure status to controls in unmatched case-control studies: alcohol use and nearly lethal suicide attempts

In case-control studies, determination of alcohol consumption by cases immediately prior to the injury event is often conceptually straightforward. However, determination of consumption status by controls is difficult because they lack a reference point, especially when cases and controls are not individually matched. We describe a method of assigning alcohol consumption status to controls using a 24-hour drinking history, the distribution in time of case events, and the random assignment of a specific time period to each control subject. This methodology offers a practical approach for determining alcohol consumption status among control subjects immediately prior to a case event, when controls lack a reference point and have not been individually matched to cases. Published in 2001 by John Wiley & Sons, Ltd.